Radiated emission measurements in open areas

In the open-field radiated emission test (OATS) measurement, it is necessary to determine whether the signal displayed on the signal analyzer is the radiated signal of the device under test or the signal from the surrounding environment. It is a cumbersome task to repeatedly turn off the device under test to check whether the signal disappears, and then turn on the device power to continue the test.

This article introduces several methods for distinguishing between the signals of the device under test and the environmental signals, hoping to solve the problems encountered in OATS measurement work. These techniques include: using a turntable to identify the signals of the device under test and the environmental signals; using dual antennas for signal identification; using shielding methods and tuning and monitoring methods to identify signals. This article also introduces how to use the trace math function to remove environmental signals from the measured signals displayed by the signal analyzer. It also discusses the use of EMI receivers and signal analyzers with signal list functions to remove environmental signals and leave only the signals of the device under test.


2. Several methods for identifying environmental signals

2.1 Using the turntable method to determine the measured signal

If the power of the device under test is not prone to frequent cycling, a turntable can be used during the measurement to help identify the signal under test. In most cases, the signal radiated from the device under test is stronger in one plane than in another. As the turntable rotates, the amplitude of the signal radiated by the device under test will change. Rotate the turntable and observe the signal amplitude at the receiver to find the maximum and minimum signals. Record the maximum value, frequency, and turntable position of the signal that is above the specification limit or margin to facilitate further analysis of the signal.

The method of measuring with a rotating turntable can identify environmental signals because the amplitude of the environmental signal does not change when the turntable rotates. If multiple measurements are made in the same area, it is recommended to use a table to record the frequency and amplitude of the environmental signal.

2.2 Dual-antenna method to identify environmental signals

A very flexible way to identify environmental signals is to use the dual antenna method. Usually the first antenna is used for normal radiation measurement and is placed 10 meters away from the plane of the device under test as defined by the specification. The second antenna is placed at twice the distance. The two antennas are connected to the receiver using an RF switch. The first antenna is used for normal measurement. When an abnormal signal is encountered, it may be an environmental signal or a radiation signal from the device under test. Switch to the second antenna and measure again. If the signal amplitudes of the two measurements are roughly the same, the signal is an environmental signal. If the amplitude value of the second measurement is about 6 dB lower, the signal is emitted by the device under test.

The antennas used in the dual antenna method should be of the same type and have similar transmission factors. In addition, the two antennas and the device under test should be placed in a straight line as much as possible to ensure that the radiation received by each antenna comes from the same plane of the device under test.

2.3 Shielding method to identify environmental signals

If it is not convenient to turn the power of the device under test on and off, a less common but economical method is to place a shield between the device under test and the measurement antenna. The shield is usually an aluminum plate, perpendicular to the antenna. It is important that the shield must be grounded and connected to the OATS ground plane. Using this method may be problematic on windy days because it is difficult to ensure that the aluminum plate is stably fixed to the test site.

2.4 Tuning and listening method to identify environmental signals

Many signal analyzers and EMI receivers offer the ability to demodulate an AM or FM signal at a marker. Place a marker at the signal point of interest and select the appropriate demodulation method such as AM, FM, or PM to demodulate. While the signal is being demodulated, the scan will pause for a while at the marker. This is very useful for quickly identifying whether it is a local radio station or a signal radiated by the device under test. This method is not suitable for identifying digital communication signals (mobile phones, digital TV, etc.).

2.5 Signal List Method to Remove Ambient Signals

One of the functions of signal analyzers and EMI receivers is to put signal peaks into a signal list. These peaks may be radiated signals from the device under test or signals from the surrounding environment. Figure 1 is a typical captured radiated signal display, which includes both the measured signal and the environmental signal.

Figure 1. Radiated emissions

Operations on the signal list, such as using the functions of marking signals and deleting signals, can reduce the signal list to only the radiation signals of the device under test and the environmental signals that did not appear in one of the two scans.

The operation process is very simple. In the first scan, the power of the device under test is not turned on, and the signal list is recorded. Then the power of the device under test is turned on, and a second scan is performed, and the signal list is recorded again. Search for repeated signals recorded in both scans. These repeated signals are environmental signals that appear in both scans. Figure 2 shows the signal list with repeated signals marked.

Figure 2. Marking repetitive signals

After deleting the duplicate signals, the remaining signals are the device under test signal and the environmental signal that did not appear in one of the two scans. Figure 3 shows the remaining signal after deleting the duplicate signals. [page]

Figure 3.   The remaining signal after deleting the duplicate signal

2.6 Removing the ambient signal from the displayed spectrum

The fundamental purpose of EMI measurement is to identify the radiated signals of the device under test and remove other irrelevant signals. Once all the signals under test can be identified, quasi-peak or average measurements can be performed to ensure that all radiated signals of the device under test can pass the limits specified in the specification. If the signal analyzer does not have the above-mentioned signal list function, you can also use the Advanced Analysis function to remove the displayed environmental signals.

The following method depends on the capabilities of modern signal analyzers, which must have enough data points to maintain the spectral resolution of the wideband scan. The more data points, the less likely it is that the signal is not acquired and the displayed amplitude is lower than expected.

The principle of eliminating the displayed ambient signal is quite simple. One trace contains both the ambient signal and the DUT signal, and the other trace contains only the ambient signal. The two traces are subtracted to obtain a signal trace with only the DUT signal.

Although the principle is simple, the actual implementation process is still relatively complicated. The ultimate goal is to obtain a set of suspicious measured signals and then re-measure them using quasi-peak detection or average detection.

The key to success is to divide and conquer. The traces in Figure 4 show that many of the measured signals and the ambient signals are very closely spaced. The frequency accuracy of the spectrum analyzer determines how closely a group of signals can be displayed on the screen. If the frequency accuracy is very low, the signals need to be spaced farther apart to be distinguishable. If the frequency accuracy is quite good, then very closely spaced signals can be distinguished. In addition, the larger the sweep width, the closer the signals appear.

Here is a brief summary of the process of removing ambient signals from the spectrum display:

1.   Set up the device under test on the turntable and set up the antenna according to the specifications.

2.   Connect the signal analyzer and antenna, load the correction factor, select the limit line required by the specification, and ensure that the sensitivity between the noise floor and the limit line is sufficient, usually at least 6 dB. Otherwise, you need to add a low noise amplifier to improve the sensitivity.

3.   Turn on the power of the device under test and observe the spectrum display. Adjust the cutoff frequency to reduce the span until the spectrum displays about 25 or 30 signals. Note that the spectrum now includes the ambient signal and the device under test signal. Increase the trace average to reduce noise. Save the spectrum as trace 1.

4.   Turn off the power of the device under test and save the spectrum as Trace 2.

5.   Use the trace add and subtract function, in this case to calculate the power difference, by subtracting trace 2 from trace 1 to get trace 3. The signal on trace 3 includes the device under test signal and any signal that appeared during the period when traces 1 and 2 were saved.

6.   Record the amplitude and corresponding frequency of the signal on trace 3 for further analysis.

7.   Rotate the turntable 90 degrees and repeat steps 3 to 6. Save all measurement results in a large table.

8.   Check the saved measurement results. If the signal amplitude is higher than the limit line, amplify the signal and perform a "measure at marker" measurement. Ensure that the quasi-peak measurement result is lower than the limit line.

2.6.1 Specific steps to eliminate environmental signals

The process of removing ambient signals uses a modern mid-range signal analyzer and can also be accomplished on an economical mid-range signal analyzer.

Shown below is a 10 meter biconical antenna that meets the measurement limits of EN55022. Continuously reducing the cutoff frequency shows only about 20 signals, including the device under test signal.

2.6.1.1 Simultaneous display of the spectrum of the ambient signal and the measured signal

Figure 4 (Trace 1) shows a spectrum that includes both the ambient signal and the signal from the device under test. Trace averaging is used to smooth the displayed trace. In this example, 100 trace averages were used to reduce the noise floor and stabilize the ambient signal. This is the trace with the ambient signal removed, leaving only the device under test signal.

Note that the ambient environment is changing as the signal fades. The signal amplitude may change over time. When the trace is stored, if the ambient signal amplitude decreases or disappears completely, then the ambient signal may appear as a DUT signal.

Figure 4: The device under test signal plus the environmental signal

Figure 5 (Trace 2, blue) shows the ambient signal when the device under test is powered off. This trace is also averaged 100 times to reduce noise and intermittent signals. Trace 2 is subtracted from Trace 1. If an ambient signal is present on Trace 2 but not on Trace 1, a negative spike will appear as a result. This negative spike is due to the subtraction.

Figure 5 Environmental signal

Subtracting trace 2 from trace 1 uses the signal analyzer's power difference calculation function. The power difference function uses the operator Operands. Define trace 1 as Operand 1 and trace 2 as Operand 2. Subtract trace 2 from trace 1 to get trace 3, which is the difference between the two. At this time, trace 1 is in static (View) mode, trace 2 is also in static mode, and trace 3 is in active (clear write) mode.

Figure 6. DUT signal and ambient signal trace minus ambient signal trace

Most of the environmental signals can be eliminated, and the amplitude values ​​of the remaining environmental signals are very low. These residual signals are generated by the change of frequency, that is, the frequency points on trace 1 and the frequency points on trace 2 have changed.

If a trace is divided into 5000 points, each point represents a bandwidth of 54 kHz. The number of points is very important for eliminating ambient signals. If too few points are selected, the DUT signal may be deleted as an ambient signal when the interval between the DUT signal and the ambient signal is too close. On the other hand, if too many points are selected, once the frequency of the ambient signal changes slightly, the points on the two traces may be different. In this way, the ambient signal point cannot be eliminated when the two traces are subtracted. This situation usually occurs at FM stations. If you look closely at trace 3, you will find that some signals show a much lower amplitude value. These signal points are the ones whose corresponding frequencies on traces 1 and 2 have changed. The amplitude of the DUT signal collected on trace 1 has almost no change or only a small change. [page]

 Figure 7 Display after turning off Trace 1 and Trace 2

With traces 1 and 2 turned off, it is easier to see the remaining signals. The DUT signal is easily identifiable. All other signals are below the limit or margin lines.

The trace in Figure 7, in some cases, will appear to extend beyond the bottom of the screen. This is because the calculated power difference is a negative value that exceeds the measurement display range. The next step is to save the results and make a final measurement of the quasi-peak or average value.

Figure 8 Display line and measured signal trace

The display line is used to collect the device under test signal. The signal above the display line is recorded in the peak list for further analysis. Usually the display line is located at the limit line or margin line. Note that not all signals on the trace come from the device under test. It may also be a signal that is not collected by trace 1 or trace 2 at the same time.

Figure 9 Peak value list of the device under test signal

Identify the signals that need to be analyzed in the next step, record them in a list and save them as a .csv file. Recall the file, use the specified detector to measure the amplitude and frequency of these signal points, record the result list and save it, which can be used for future measurements. If there are signals that do not meet the specifications, the device under test needs to be redesigned and re-evaluated.

3. Comparison of different identification methods of environmental signals

3.1 Turntable method

When it is difficult to turn off the power of the device under test, the turntable method is suitable. It is necessary to record and compare the data measured from four directions to distinguish the radiation signal of the device under test.

 
3.2 Dual Antenna Method

This method is the fastest for single signal identification. By simply switching between the two antennas, you can easily see whether the signal is an environmental signal or a signal from the device under test. However, this method requires a very large test site.

3.3 Shielding method

Setting up a shield is the lowest cost method. However, this method is a bit cumbersome and requires installation and removal. When there are multiple signals to be analyzed, the shield method is suitable.

3.4 Signal List Method

The signal list method is very fast and allows the user to directly select the radiated signal of the device under test in the signal list for the final measurement. The signal analyzer used must have a signal list function, such as a signal analyzer installed with the Agilent N6141A EMC measurement application software.

4       Conclusion

The method used to remove environmental signals depends on the device under test, the surrounding environment, and the site of use. For example: the device under test is relatively complex (i.e.:If the environment interference is relatively small (far below the limit line), the turntable method can be used.